Title :
Fault detection in a mixed H2/H∞ setting
Author :
Khosrowjerdi, Mohammad-Javad ; Nikoukhah, Ramine ; Safari-Shad, Nader
Author_Institution :
Dept. of Electr. Eng., K.N. Toosi Technol. Univ., Tehran, Iran
Abstract :
In this paper, we study the problem of fault detection in linear time-invariant systems, which contain two types of inputs: inputs with fixed and known spectral densities and inputs with bounded power. It is shown that this problem can be formulated as a mixed H2/H∞ filtering problem where the filter gain is computed by solving a pair of coupled Riccati equations. Alternatively, it is shown that the filter gain may be computed by transforming the problem into a convex optimization problem with a set of linear matrix inequality (LMI) constraints. The application of our results is illustrated by an example.
Keywords :
H∞ control; Riccati equations; convex programming; fault location; filtering theory; linear matrix inequalities; linear systems; LMI; Riccati equations; convex optimization problem; fault detection; filter gain; linear matrix inequality; linear time-invariant systems; mixed H2/H∞ setting; spectral densities; Constraint optimization; Electric breakdown; Electrical fault detection; Fault detection; Filtering; Linear matrix inequalities; Nonlinear filters; Reliability theory; Riccati equations; Safety;
Conference_Titel :
Decision and Control, 2003. Proceedings. 42nd IEEE Conference on
Print_ISBN :
0-7803-7924-1
DOI :
10.1109/CDC.2003.1272817